Add lang alumni

[dtolnay]

Citations:

lang-shepherds
rust-lang/prev.rust-lang.org#823. There are 4 other former members of lang-shepherds: Centril, cramertj, joshtriplett, scottmcm. I have not included them in "lang-shepherds alumni" because they are current lang team members or alumni. The precedent in rust-lang/prev.rust-lang.org#914, rust-lang/prev.rust-lang.org#1072, rust-lang/prev.rust-lang.org#1176 is that when a member of lang-shepherds joins the lang team, it atomically subsumes their lang-shepherds role. Lang-shepherds is not listed on the website (because it is an archived team) but if it were, it would not make sense to list a current lang team member under lang-shepherds saying "We also want to thank all past members for their invaluable contributions" when that person is effectively still performing a superset of that role.

reference
#307

lang
rust-lang/prev.rust-lang.org#414

style
rust-lang/prev.rust-lang.org#523

opsem
Not touched in this PR. But I wonder if it might make sense to list members of the erstwhile wg-unsafe-code-guidelines as "alumni" of opsem. In #926, opsem subsumed wg-unsafe-code-guidelines, including renaming the t-lang/unsafe-code-guidelines Zulip stream to t-opsem. It might be fair for wg-unsafe-code-guidelines members to be listed in an alumni section on https://www.rust-lang.org/governance/teams/lang#opsem ("We also want to thank all past members for their invaluable contributions"). @JakobDegen? This would apply to avadacatavra, comex, nikomatsakis. The counterargument is that sometimes wg membership has been laxer than team membership, where depending on the working group (I am not familiar with this one), anyone who shows up gets their name on the roster. It's not necessarily fair to get credited with team alumni status.

types
Similar situation to opsem. When types team was created in #760, wg-traits was archived saying "this has effectively been replaced by the types team" and "we want to archive this to recognize the various people that contributed while the group was under the 'wg-traits' label" implying some continuity of "the group" under a different name and possibly more formalized membership process. See also #858.

[dtolnay]

The counterargument is that sometimes wg membership has been laxer than team membership, where depending on the working group (I am not familiar with this one), anyone who shows up gets their name on the roster. It's not necessarily fair to get credited with team alumni status.

To expand: this perception is informed by #992 and #656 where there is a distinction between "has ever been a member" vs "has contributed enough to be recognized as alumni".

My impression is that the typical case for a team (as opposed to typical working group) is someone added as a member has probably already contributed in that area enough to be worth recognizing regardless of what more they might contribute during their membership. Backfilling alumni from git history makes sense: for example the glaring omission of eddyb from lang alumni. Whereas there are working groups for which becoming a member is one of the first steps to contributing in the area; this makes trying to backfill wg alumni from git history not necessarily wise. Ditto when a working group evolves into a team.

[petrochenkov]

someone added as a member has probably already contributed in that area enough

In a few cases (triage team? prioritization team? can't find the relevant links right now) people were added just because they answered an advertisement on IRLO, but then they never actually did anything.

[dtolnay]

Right; both of those are working groups.

team/teams/wg-triage.toml

Lines 1 to 3 in 59784db

	name = "wg-triage"
	subteam-of = "release"
	kind = "working-group"

team/teams/wg-prioritization.toml

Lines 1 to 3 in 59784db

	name = "wg-prioritization"
	subteam-of = "compiler"
	kind = "working-group"

[jackh726]

implying some continuity of "the group" under a different name and possibly more formalized membership process

Yes, the membership process for the types team is much more formalized than for the wg (it's also true that wg-traits was not under lang anyways). Really, the team subsumes the responsibilities of the former wg, but is different. I think it's appropriate to keep the wg-traits archived separately :)